WebOS, PlanetLab, Starfish - Revision history

Jmahonin: /* Learning from PlanetLab */

2008-03-19T19:55:49Z

Learning from PlanetLab

← Older revision		Revision as of 19:55, 19 March 2008
Line 145:		Line 145:
	* Never had a good API, inconsistent, ever changing, unstable programming environment		* Never had a good API, inconsistent, ever changing, unstable programming environment

	=== Focus on the machine room		=== Focus on the machine room ===
	==== Pro ====		==== Pro ====
	* Allocate big machine here, other there, etc...		* Allocate big machine here, other there, etc...
	==== Con ====		==== Con ====
	* Bad for distributed OSes		* Bad for distributed OSes

2008-03-19T19:55:11Z

Learning from PlanetLab

@@ Line 130: / Line 130: @@
 ==== Con ====
 * Weak isolation between experiments
 * Having a homogenous test-bed is poor for distributed experimentation

2008-03-19T19:50:08Z

Learning from PlanetLab

← Older revision		Revision as of 19:50, 19 March 2008
Line 120:		Line 120:

	=== Provide only best-effort service ===		=== Provide only best-effort service ===
	=== Pro ===		==== Pro ====
	* No limit on the number of processes which can be run		* No limit on the number of processes which can be run
	=== Con ===		==== Con ====
	* Other processes may crowd out computation of other processes		* Other processes may crowd out computation of other processes (cpu/disk hogging)

			=== Linux is the execution environment ===
			==== Pro ====
			* Provides a familiar programming environment
			==== Con ====
			* Weak isolation between experiments
			* Having a homogenous test-bed is poor for distributed experimentation
			*

2008-03-19T19:42:11Z

Learning from PlanetLab

← Older revision		Revision as of 19:42, 19 March 2008
Line 110:		Line 110:
	==== Pro ====		==== Pro ====
	* Provide bare-bones management, try to foster competition between 3rd party services		* Provide bare-bones management, try to foster competition between 3rd party services
	==== Con =====		==== Con ====
	* No motivation to do so, it's hard work		* No motivation to do so, it's hard work

Line 116:		Line 116:
	==== Pro ====		==== Pro ====
	* Free bandwidth!		* Free bandwidth!
	==== Con =====		==== Con ====
	* No incentive to conserve bandwidth		* No incentive to conserve bandwidth

	=== Provide only best-effort service ===		=== Provide only best-effort service ===
	=== Pro ===		=== Pro ===
			* No limit on the number of processes which can be run
	=== Con ===		=== Con ===
			* Other processes may crowd out computation of other processes

2008-03-19T19:40:05Z

PlanetLab Learning

@@ Line 94: / Line 94: @@
 * Bandwidth allocation was slice-based, not node based
-== PlanetLab Learning ==
+== Learning from PlanetLab==
-=== Bar ===
+=== Centralized Trusts ===

2008-03-19T19:27:27Z

Experiences Building PlanetLab

← Older revision		Revision as of 19:27, 19 March 2008
Line 85:		Line 85:
	* Implementing a trust relationship between node owners and service developers (users)		* Implementing a trust relationship between node owners and service developers (users)
	=== PLC ===		=== PLC ===
			==== Summary ====
	* Fulfills the trust mechanism required for the network		* Fulfills the trust mechanism required for the network
	* Acts as a middle man / mediator between node owners and users		* Acts as a middle man / mediator between node owners and users
	* If someone breaks into the PLC however, entire system is compromised		* If someone breaks into the PLC however, entire system is compromised
			==== Implementation ====
			* Nodes are split into slices using VServers, lightweight process groups
			==== Criticisms ====
			* Bandwidth allocation was slice-based, not node based

	== PlanetLab Learning ==		== PlanetLab Learning ==
	=== Bar ===		=== Bar ===

2008-03-19T19:13:31Z

PlanetLab

← Older revision		Revision as of 19:13, 19 March 2008
Line 75:		Line 75:


	== PlanetLab ==		== Experiences Building PlanetLab ==
	=== ~~Foo~~ ===		=== Key ideas ===
			* Global platform to distribute/test network services
			* Scale
			* Should be able to monitor and stop disruptive traffic
			=== Challenges ===
			* Resource allocation is significant, especially in a distributed system such as PlanetLab
			* Providing a global platform for long-term services
			* Implementing a trust relationship between node owners and service developers (users)
			=== PLC ===
			* Fulfills the trust mechanism required for the network
			* Acts as a middle man / mediator between node owners and users
			* If someone breaks into the PLC however, entire system is compromised

	== PlanetLab Learning ==		== PlanetLab Learning ==
	=== Bar ===		=== Bar ===

2008-03-19T19:03:29Z

← Older revision		Revision as of 19:03, 19 March 2008
Line 73:		Line 73:
	* Had a Java interface which allowed any node to login and see full system status		* Had a Java interface which allowed any node to login and see full system status
	* Used the distributed nature of daemons to communicate this information		* Used the distributed nature of daemons to communicate this information


			== PlanetLab ==
			=== Foo ===

			== PlanetLab Learning ==
			=== Bar ===

2008-03-19T19:00:29Z

Starfish

← Older revision		Revision as of 19:00, 19 March 2008
Line 70:		Line 70:
	* Provides process migration		* Provides process migration
	* Maybe better suited to be implemented in the application, not OS?		* Maybe better suited to be implemented in the application, not OS?
			==== Management ====
			* Had a Java interface which allowed any node to login and see full system status
			* Used the distributed nature of daemons to communicate this information

2008-03-19T18:56:46Z

Starfish

← Older revision		Revision as of 18:56, 19 March 2008
Line 66:		Line 66:
	* Slow performance?		* Slow performance?
	* Not entirely portable, maybe Just-In-Time?		* Not entirely portable, maybe Just-In-Time?
			==== Checkpointing ====
			* Save / resume state
			* Provides process migration
			* Maybe better suited to be implemented in the application, not OS?